Issue with SEP and prediction intervals for monthly/annual aggregate analyses

[gfkltn]

Incorrect SEPs and prediction intervals can be computed when using unit values to develop regression equations for flux and then predicting monthly or annual average fluxes. As illustrated in the output of monthly average fluxes computed from a unit-value model, the standard errors are sometimes, but not always, larger than the SEPs. That should never happen. As far as I have observed, this issue can occur with monthly and annual flux estimates. I have not seen similar problems with daily flux estimates.